Guilherme H. Resende, Luiz F. Nery, Fabrício Benevenuto, Savvas Zannettou, Flavio Figueiredo
Language is a dynamic aspect of our culture that changes when expressed in different technologies/communities. Online social networks have enabled the diffusion and evolution of different dialects, including African American English (AAE). However, this increased usage is not without barriers. One particular barrier is how sentiment (Vader, TextBlob, and Flair) and toxicity (Google's Perspective and the open-source Detoxify) methods present biases towards utterances with AAE expressions. Consider Google's Perspective to understand bias. Here, an utterance such as ``All n*ggers deserve to die respectfully. The police murder us.'' it reaches a higher toxicity than ``African-Americans deserve to die respectfully. The police murder us.''. This score difference likely arises because the tool cannot understand the re-appropriation of the term ``n*gger''. One explanation for this bias is that AI models are trained on limited datasets, and using such a term in training data is more likely to appear in a toxic utterance. While this may be plausible, the tool will make mistakes regardless. Here, we study bias on two Web-based (YouTube and Twitter) datasets and two spoken English datasets. Our analysis shows how most models present biases towards AAE in most settings. We isolate the impact of AAE expression usage via linguistic control features from the Linguistic Inquiry and Word Count (LIWC) software, grammatical control features extracted via Part-of-Speech (PoS) tagging from Natural Language Processing (NLP) models, and the semantic of utterances by comparing sentence embeddings from recent language models. We present consistent results on how a heavy usage of AAE expressions may cause the speaker to be considered substantially more toxic, even when speaking about nearly the same subject. Our study complements similar analyses focusing on small datasets and/or one method only.
Flavio Figueiredo, Jussara M. Almeida, Marcos André Gonçalves, Fabrício Benevenuto
We here focus on the problem of predicting the popularity trend of user generated content (UGC) as early as possible. Taking YouTube videos as case study, we propose a novel two-step learning approach that: (1) extracts popularity trends from previously uploaded objects, and (2) predicts trends for new content. Unlike previous work, our solution explicitly addresses the inherent tradeoff between prediction accuracy and remaining interest in the content after prediction, solving it on a per-object basis. Our experimental results show great improvements of our solution over alternatives, and its applicability to improve the accuracy of state-of-the-art popularity prediction methods.
Flavio Figueiredo, Jussara M. Almeida, Yasuko Matsubara, Bruno Ribeiro, Christos Faloutsos
How many listens will an artist receive on a online radio? How about plays on a YouTube video? How many of these visits are new or returning users? Modeling and mining popularity dynamics of social activity has important implications for researchers, content creators and providers. We here investigate the effect of revisits (successive visits from a single user) on content popularity. Using four datasets of social activity, with up to tens of millions media objects (e.g., YouTube videos, Twitter hashtags or LastFM artists), we show the effect of revisits in the popularity evolution of such objects. Secondly, we propose the Phoenix-R model which captures the popularity dynamics of individual objects. Phoenix-R has the desired properties of being: (1) parsimonious, being based on the minimum description length principle, and achieving lower root mean squared error than state-of-the-art baselines; (2) applicable, the model is effective for predicting future popularity values of objects.
Gabriel Souza, Flavio Figueiredo, Alexei Machado, Deborah Guimarães
In recent years, deep learning has achieved formidable results in creative computing. When it comes to music, one viable model for music generation are Transformer based models. However, while transformers models are popular for music generation, they often rely on annotated structural information. In this work, we inquire if the off-the-shelf Music Transformer models perform just as well on structural similarity metrics using only unannotated MIDI information. We show that a slight tweak to the most common representation yields small but significant improvements. We also advocate that searching for better unannotated musical representations is more cost-effective than producing large amounts of curated and annotated data.
Flavio Figueiredo, José Geraldo Fernandes, Jackson Silva, Renato M. Assunção
Copulas are powerful statistical tools for capturing dependencies across data dimensions. Applying Copulas involves estimating independent marginals, a straightforward task, followed by the much more challenging task of determining a single copulating function, $C$, that links these marginals. For bivariate data, a copula takes the form of a two-increasing function $C: (u,v)\in \mathbb{I}^2 \rightarrow \mathbb{I}$, where $\mathbb{I} = [0, 1]$. This paper proposes 2-Cats, a Neural Network (NN) model that learns two-dimensional Copulas without relying on specific Copula families (e.g., Archimedean). Furthermore, via both theoretical properties of the model and a Lagrangian training approach, we show that 2-Cats meets the desiderata of Copula properties. Moreover, inspired by the literature on Physics-Informed Neural Networks and Sobolev Training, we further extend our training strategy to learn not only the output of a Copula but also its derivatives. Our proposed method exhibits superior performance compared to the state-of-the-art across various datasets while respecting (provably for most and approximately for a single other) properties of C.
Flavio Figueiredo, Tales Panoutsos, Nazareno Andrade
Analyzing musical influence networks, such as those formed by artist influence or sampling, has provided valuable insights into contemporary Western music. Here, computational methods like centrality rankings help identify influential artists. However, little attention has been given to how influence changes over time. In this paper, we apply Bayesian Surprise to track the evolution of musical influence networks. Using two networks -- one of artist influence and another of covers, remixes, and samples -- our results reveal significant periods of change in network structure. Additionally, we demonstrate that Bayesian Surprise is a flexible framework for testing various hypotheses on network evolution with real-world data.
Flavio Figueiredo, Guilherme Borges, Pedro O. S. Vaz de Melo, Renato M. Assunção
We here focus on the task of learning Granger causality matrices for multivariate point processes. In order to accomplish this task, our work is the first to explore the use of Wold processes. By doing so, we are able to develop asymptotically fast MCMC learning algorithms. With $N$ being the total number of events and $K$ the number of processes, our learning algorithm has a $O(N(\,\log(N)\,+\,\log(K)))$ cost per iteration. This is much faster than the $O(N^3\,K^2)$ or $O(K^3)$ for the state of the art. Our approach, called GrangerBusca, is validated on nine datasets. This is an advance in relation to most prior efforts which focus mostly on subsets of the Memetracker data. Regarding accuracy, GrangerBusca is three times more accurate (in Precision@10) than the state of the art for the commonly explored subsets Memetracker. Due to GrangerBusca's much lower training complexity, our approach is the only one able to train models for larger, full, sets of data.
Mariana Arantes, Flavio Figueiredo, Jussara M. Almeida
Faced with the challenge of attracting user attention and revenue, social media websites have turned to video advertisements (video-ads). While in traditional media the video-ad market is mostly based on an interaction between content providers and marketers, the use of video-ads in social media has enabled a more complex interaction, that also includes content creator and viewer preferences. To better understand this novel setting, we present the first data-driven analysis of video-ad exhibitions on YouTube.
Flavio Figueiredo, Giovanni Martinelli, Henrique Sousa, Pedro Rodrigues, Frederico Pedrosa, Lucas N. Ferreira
Recent advances in AI music (AIM) generation services are currently transforming the music industry. Given these advances, understanding how humans perceive AIM is crucial both to educate users on identifying AIM songs, and, conversely, to improve current models. We present results from a listener-focused experiment aimed at understanding how humans perceive AIM. In a blind, Turing-like test, participants were asked to distinguish, from a pair, the AIM and human-made song. We contrast with other studies by utilizing a randomized controlled crossover trial that controls for pairwise similarity and allows for a causal interpretation. We are also the first study to employ a novel, author-uncontrolled dataset of AIM songs from real-world usage of commercial models (i.e., Suno). We establish that listeners' reliability in distinguishing AIM causally increases when pairs are similar. Lastly, we conduct a mixed-methods content analysis of listeners' free-form feedback, revealing a focus on vocal and technical cues in their judgments.
Breno Matos, Francisco Galuppo, Rennan Cordeiro, Flavio Figueiredo
With over a billion active users, TikTok's video-sharing service is currently one of the largest social media websites. This rise in TikTok's popularity has made the website a central platform for music discovery. In this paper, we analyze how TikTok helps to revitalize older songs. To do so, we use both the popularity of songs shared on TikTok and how the platform allows songs to propagate to other places on the Web. We analyze data from TokBoard, a website measuring such popularity over time, and Google Trends, which captures songs' overall Web search interest. Our analysis initially focuses on whether TokBoard can cause (Granger Causality) popularity on Google Trends. Next, we examine whether TikTok and Google Trends share the same virality patterns (via a Bass Model). To our knowledge, we are one of the first works to study song re-popularization via TikTok.
Flavio Figueiredo, Marcos André Gonçalves, Jussara M. Almeida
We here present a simple and effective model to predict the popularity of web content. Our solution, which is the winner of two of the three tasks of the ECML/PKDD 2014 Predictive Analytics Challenge, aims at predicting user engagement metrics, such as number of visits and social network engagement, that a web page will achieve 48 hours after its upload, using only information available in the first hour after upload. Our model is based on two steps. We first use time series clustering techniques to extract common temporal trends of content popularity. Next, we use linear regression models, exploiting as predictors both content features (e.g., numbers of visits and mentions on online social networks) and metrics that capture the distance between the popularity time series to the trends extracted in the first step. We discuss why this model is effective and show its gains over state of the art alternatives.
Guilherme Soares S. dos Santos, Flavio Figueiredo
Music digitalization has introduced new forms of composition known as "musical borrowings", where composers use elements of existing songs -- such as melodies, lyrics, or beats -- to create new songs. Using Who Sampled data and Google Trends, we examine how the popularity of a borrowing song affects the original. Employing Regression Discontinuity Design (RDD) for short-term effects and Granger Causality for long-term impacts, we find evidence of causal popularity boosts in some cases. Borrowee songs can revive interest in older tracks, underscoring economic dynamics that may support fairer compensation in the music industry.
Flavio Figueiredo, Jussara M. Almeida, Marcos André Gonçalves, Fabrício Benevenuto
Understanding the factors that impact the popularity dynamics of social media can drive the design of effective information services, besides providing valuable insights to content generators and online advertisers. Taking YouTube as case study, we analyze how video popularity evolves since upload, extracting popularity trends that characterize groups of videos. We also analyze the referrers that lead users to videos, correlating them, features of the video and early popularity measures with the popularity trend and total observed popularity the video will experience. Our findings provide fundamental knowledge about popularity dynamics and its implications for services such as advertising and search.
Flavio Figueiredo, Bruno Ribeiro, Jussara Almeida, Christos Faloutsos
Which song will Smith listen to next? Which restaurant will Alice go to tomorrow? Which product will John click next? These applications have in common the prediction of user trajectories that are in a constant state of flux over a hidden network (e.g. website links, geographic location). What users are doing now may be unrelated to what they will be doing in an hour from now. Mindful of these challenges we propose TribeFlow, a method designed to cope with the complex challenges of learning personalized predictive models of non-stationary, transient, and time-heterogeneous user trajectories. TribeFlow is a general method that can perform next product recommendation, next song recommendation, next location prediction, and general arbitrary-length user trajectory prediction without domain-specific knowledge. TribeFlow is more accurate and up to 413x faster than top competitors.
Felipe Giori, Flavio Figueiredo
Pairwise Causal Discovery is the task of determining causal, anticausal, confounded or independence relationships from pairs of variables. Over the last few years, this challenging task has promoted not only the discovery of novel machine learning models aimed at solving the task, but also discussions on how learning the causal direction of variables may benefit machine learning overall. In this paper, we show that Quantitative Information Flow (QIF), a measure usually employed for measuring leakages of information from a system to an attacker, shows promising results as features for the task. In particular, experiments with real-world datasets indicate that QIF is statistically tied to the state of the art. Our initial results motivate further inquiries on how QIF relates to causality and what are its limitations.
Elizeu Santos-Neto, Flavio Figueiredo, Nigini Oliveira, Nazareno Andrade, Jussara Almeida, Matei Ripeanu
Tagging is a popular feature that supports several collaborative tasks, including search, as tags produced by one user can help others finding relevant content. However, task performance depends on the existence of 'good' tags. A first step towards creating incentives for users to produce 'good' tags is the quantification of their value in the first place. This work fills this gap by combining qualitative and quantitative research methods. In particular, using contextual interviews, we first determine aspects that influence users' perception of tags' value for exploratory search. Next, we formalize some of the identified aspects and propose an information-theoretical method with provable properties that quantifies the two most important aspects (according to the qualitative analysis) that influence the perception of tag value: the ability of a tag to reduce the search space while retrieving relevant items to the user. The evaluation on real data shows that our method is accurate: tags that users consider more important have higher value than tags users have not expressed interest.
Tales Panoutsos, Rodrygo L. T. Santos, Flavio Figueiredo
In this paper, we introduce Symmetric Low-Rank Adapters, an optimized variant of LoRA with even fewer weights. This method utilizes Low-Rank Symmetric Weight Matrices to learn downstream tasks more efficiently. Traditional LoRA accumulates fine-tuning weights with the original pre-trained weights via a Singular Value Decomposition (SVD) like approach, i.e., model weights are fine-tuned via updates of the form $BA$ (where $B \in \mathbb{R}^{n\times r}$, $A \in \mathbb{R}^{r\times n}$, and $r$ is the rank of the merged weight matrix). In contrast, our approach, named SymLoRA, represents fine-tuning weights as a Spectral Decomposition, i.e., $Q \, diag(Λ)\, Q^T$, where $Q \in \mathbb{R}^{n\times r}$ and $Λ\in \mathbb{R}^r$. SymLoRA requires approximately half of the finetuning weights. Here, we show that this approach has negligible losses in downstream efficacy.
Renato Assunção, Flávio Figueiredo, Francisco N. Tinoco Júnior, Léo M. de Sá-Freire, Fábio Silva
A fundamental task in statistical learning is quantifying the joint dependence or association between two continuous random variables. We introduce a novel, fully non-parametric measure that assesses the degree of association between continuous variables $X$ and $Y$, capable of capturing a wide range of relationships, including non-functional ones. A key advantage of this measure is its interpretability: it quantifies the expected relative loss in predictive accuracy when the distribution of $X$ is ignored in predicting $Y$. This measure is bounded within the interval [0,1] and is equal to zero if and only if $X$ and $Y$ are independent. We evaluate the performance of our measure on over 90,000 real and synthetic datasets, benchmarking it against leading alternatives. Our results demonstrate that the proposed measure provides valuable insights into underlying relationships, particularly in cases where existing methods fail to capture important dependencies.
Martino Trevisan, Luca Vassio, Idilio Drago, Marco Mellia, Fabricio Murai, Flavio Figueiredo, Ana Paula Couto da Silva, Jussara M. Almeida
Online Social Networks (OSNs) allow personalities and companies to communicate directly with the public, bypassing filters of traditional medias. As people rely on OSNs to stay up-to-date, the political debate has moved online too. We witness the sudden explosion of harsh political debates and the dissemination of rumours in OSNs. Identifying such behaviour requires a deep understanding on how people interact via OSNs during political debates. We present a preliminary study of interactions in a popular OSN, namely Instagram. We take Italy as a case study in the period before the 2019 European Elections. We observe the activity of top Italian Instagram profiles in different categories: politics, music, sport and show. We record their posts for more than two months, tracking "likes" and comments from users. Results suggest that profiles of politicians attract markedly different interactions than other categories. People tend to comment more, with longer comments, debating for longer time, with a large number of replies, most of which are not explicitly solicited. Moreover, comments tend to come from a small group of very active users. Finally, we witness substantial differences when comparing profiles of different parties.